VynarisGet your API key

Reference

LLM API pricing, in one honest table

Provider list prices, as of 2026-0730models, sorted cheapest to most expensive by input price. Rows we couldn’t verify against an official pricing page are marked .

ModelOrgWeightsContext$ in / MTok$ out / MTokWhat it’s actually good at
Command R7BCohereclosed128K$0.0375$0.1500ultra-cheap enterprise RAG and classification at scale
Llama 3.1 8BMetaopen128K$0.0500$0.1000cheap, fast, easily fine-tuned baseline for narrow tasks
Qwen3 32BAlibabaopen131K (via YaRN)$0.0800$0.2800strong reasoning-to-cost ratio in a single self-hostable dense model
Mistral Small 3Mistral AIopen128K$0.0800$0.2000efficient general chat and instruction-following at a self-hostable size
Llama 4 ScoutMetaopen10M (advertised; effective recall shorter)$0.1000$0.3000extremely long-context ingestion of codebases and long documents at low cost
GPT-4o miniOpenAIclosed128K$0.1500$0.6000cheap, fast, good-enough responses for high-volume simple tasks
Command RCohereclosed128K$0.1500$0.6000efficient RAG and enterprise search-grounded generation
Llama 4 MaverickMetaopen1M$0.2000$0.8500strong general chat and multimodal reasoning at open-weight cost
Qwen3 235B-A22BAlibabaopen131K (via YaRN)$0.2000$0.6000flagship open-weight reasoning, coding, and agentic tool use with a toggleable thinking mode
Claude 3 HaikuAnthropicclosed200K$0.2500$1.25legacy cheap-tier chat and summarization
DeepSeek V3DeepSeekopen128K$0.2700$1.10near-frontier general reasoning and coding at a fraction of closed-model cost
Gemini 2.5 FlashGoogleclosed1M$0.3000$2.50high-throughput multimodal tasks needing large context at low cost
Llama 3.3 70BMetaopen128K$0.3000$0.6000close-to-405B quality at 70B cost, a popular self-hosted workhorse
Qwen2.5 72BAlibabaopen128K$0.3500$0.4000strong general and coding performance, still widely deployed
GLM-4.5Zhipu AIopen128K$0.4000$1.60strong open-weight agentic coding and tool-use performance
DeepSeek R1DeepSeekopen128K$0.5500$2.19step-by-step reasoning, math, and coding rivaling closed reasoning models at open-weight cost
Kimi K2Moonshot AIopen128K-262K$0.5500$2.20frontier-competitive open-weight coding and agentic tool use at trillion-parameter scale
Claude Haiku 4.5Anthropicclosed200K$1.00$5.00low-latency chat, classification, and simple extraction at scale
o4-miniOpenAIclosed200K$1.10$4.40reasoning tasks at a fraction of o3's cost
Gemini 2.5 ProGoogleclosed1M$1.25$10.00native multimodal reasoning and huge-context document or video analysis
GPT-4.1OpenAIclosed1M$2.00$8.00long-context comprehension and more reliable instruction following than GPT-4o
o3OpenAIclosed200K$2.00$8.00hard multi-step math, science, and coding reasoning
Mistral Large 2Mistral AIclosed128K$2.00$6.00strong multilingual reasoning and coding, a European-hosted alternative to US frontier models
GPT-4oOpenAIclosed128K$2.50$10.00strong general-purpose chat, vision, and tool use with a good speed/quality balance
Command R+Cohereclosed128K$2.50$10.00enterprise RAG, citations, and multi-step tool use with strong grounding
Llama 3.1 405BMetaopen128K$2.70$2.70frontier-adjacent open-weight reasoning and generation quality
Claude Sonnet 4.xAnthropicclosed200K$3.00$15.00the default high-quality coding and reasoning workhorse most teams reach for first
Claude 3.5 SonnetAnthropicclosed200K$3.00$15.00strong coding and agentic tool use, the model that popularized computer use
Grok 4xAIclosed256K$3.00$15.00competitive frontier reasoning and coding, positioned against GPT-5 and Claude Opus
Claude Opus 4.xAnthropicclosed200K$15.00$75.00deepest reasoning, careful long-form writing, and complex agentic coding

18 of 30rows carry prices the source research couldn’t cross-confirm on an official provider pricing page as of 2026-07 — treat them as directional, not contractual. Price class (frontier, mid, budget, ultra-budget) is derived from list input price, not a vendor label. No benchmark scores are shown here because none of the sentences above are backed by a benchmark we ran ourselves — they’re read off the model card, not measured.

What routing does to these numbers

Every price above is what you pay if every request hits that exact model. A router instead sends each request to the cheapest model that can still do the job, so most traffic never touches the frontier price at all. Here’s one worked example, computed from the table above — not a benchmark, just arithmetic.

500,000 requests/month, averaging 600 input and 400 output tokens each, asked for at Claude Sonnet 4.x list price, served instead by DeepSeek V3 — a right-sized model in a similar capability band.

Claude Sonnet 4.x at list price
$3,900/mo
DeepSeek V3 serving the same volume
$301.00/mo
Delta
92%

This is one example at one volume, not a promise about your traffic — different prompts, task mixes, and quality bars change the answer. Run your own numbers on the cost calculator, or get an API key and see the routed price on your own requests. Full docs at /docs.